The Bullwhip Effect in Different Manufacturing Paradigm: An...

The Bullwhip Effect in Different Manufacturing Paradigm: An Analysis

Shamila Nabi KHAN1 Mohammad Ajmal KHAN2

Ramsha SOHAIL

ABSTRACT

Slight fluctuation in the customer demand produces huge fluctuations in the inventory and

production levels of the upstream member of the supply chain. This is known as demand

amplification or the bullwhip effect. The bullwhip effect produces huge inefficiencies along

the supply chain, captures much-needed working capital by high inventory levels and

results in dissatisfied customers. Analyzing a set of nine case studies, results show that

agile manufacturing is best suited to deal with a turbulent market with unpredictable

demand.

KEYWORDS: Bullwhip, Agile, Lean, Agility, Traditional, Manufacuring paradigm.

JEL CLASSIFICATION M11

INTRODUCTION

Competitive advantage can be gained through an effective supply chain management

resulting in efficiency (Chase et al, 2001). The coordination in the supply chain comprising

of manufacturers, distributors, wholesalers and retailers depends on the flow of information

(Lee et al., 1997a; Rudberg, 2003) However, distortion in the information flow leads to

supply-demand imbalance leading to a phenomenon called the ―bullwhip effect‖ (Lee et al,

1997a).

Companies rely on the customer demand to make production forecasts. In the demand-

ridden supply chain, variation in the supply-demand is often a cause of concern for the

managers. These fluctuations in the demand vary significantly between industries. In a

relatively consistent demand product it might be expected that the supply chain is more

accurate and efficient compared to products with seasonal demand and various business

cycles. However, as one moves up the supply chain from retailers to distributors the

perceived demand is amplified, which thus produces the bullwhip effect. This effect is more

present in products with stationary demand as found by the logistics executive at Proctor &

Gamble (Forrester, 1961). The same phenomenon is also found in various products and in

different industries.

1 Teaching Fellow (Ph.D. Scholar), Lahore School of Economics. Address all correspondence to

[email protected]. 2 Head of Operations Department, Dawakhana Hakim Ajmal Khan, Lahore, Pakistan.

[email protected].

Economia. Seria Management Volume 17, Issue 1, 2014

141

Various researches suggest the presence of the bullwhip effect in the supply chain of the

automobile industry. Blanchard (1983) concluded that the variance of demand was higher

that the variance of sales. Cachon (2007) also shows the presence of bullwhip effect in the

U.S automobile industry.

The objectives of this project were to build a deeper understanding of the bullwhip effect

through analyzing different case studies that have already taken place. The project sets out

to differentiate qualitatively between the demand amplification effects in various

manufacturing systems.

This paper contributes heavily in accessing the bullwhip effect in different paradigms.

There is lack of literature on bullwhip effect due to which this qualitative analysis uses

cases from different industries and different manufacturing paradigms such as traditional,

lean and agile in assessing the research question. Studies conducted on the bullwhip effect

are superficial at best. Bullwhip effect in leagile strategy is hardly documented. The paper

connected the three paradigms and assesses the bullwhip effect in leagile strategies. The

aim of this paper, thus, is to show the difference of intensity of the bullwhip effect within

mass, lean and agile manufacturing approaches. This study analyzes, through secondary

research, the existence of bullwhip effect in different manufacturing systems. First, the

three manufacturing paradigms prevalent today were categorized as: Traditional

manufacturing, lean manufacturing and agile manufacturing.

This categorization was made after looking at the way production and supply chain was

managed at these companies. The companies (or supply chains) belonged to different

industries from automotive components to food retail. Some supply chains were

international, however, most were local. All of the supply chains were a basic three-tier

network as shown in figure 1.

Figure 1. A simple three tier supply chain Source: author

Section 2 analyzes the literature behind the ideas employed. It also emphasizes how

different approaches have been developed, to reduce the demand amplification problem and

how these approaches have evolved over time in a high competitive market economy.

Section 3 gives an explanation about the methodology used to gather data for a deeper

understanding of the phenomena. Case studies were gathered from different journals.

Section 4 critically analyzes the case studies. The nine case studies are discussed and their

results analyzed in order to gain a deeper understanding of the bullwhip phenomenon.

Section 5 concludes the report with recommendations for managing the bullwhip.

1. LITERATURE REVIEW

Lee et al. (1997a), describes his idea of the bullwhip effect as the ‗the phenomenon where

orders to the supplier tend to have larger variance than sales to the buyer (i.e., demand

distortion), and the distortion propagates upstream in an amplified form (i.e., variance

amplification).‘ The name bullwhip is suggestive of the movement of the bullwhip. Small

movement at the handle of a bullwhip can create bigger or larger swing at the other end

Manufacturer Distribution Centre Retailer

Shamila Nabi KHAN, Mohammad Ajmal KHAN, Ramsha SOHAIL

142

which describes how information on demand-side becomes increasingly exaggerated and

distorted as it moves up the supply chain from customer, manufacturer towards supplier,

driving costs up and creating inefficiency. The Beer game created by the Sloan School of

Management demonstrates important principles in the supply chain. Sterman (1989) states

candidates in the game are appointed to act like the members of the supply chain; the

customer, retailer, manufacturer and the supplier. Each makes decision on their demand

order from the orders they receive unaware of the actual customer order. As expected the

inventory increases; the customer order are far lower then the aggregate inventory of the

members.

The presence of the bullwhip was of systemic nature. The parts that made up the system did

not know of its existence. It was actually an evidence of chaos in social systems.

As the demand orders move up the supply chain, the information is amplified. At the sales

end, customers order more in order to eliminate uncertainty. Hence, the retailer observes

lesser fluctuations in orders then compared to the manufacturer‘s order. The wholesalers to

manufacturer orders are exaggerated creating higher variability, which amplifies even more

when factory orders are placed to the suppliers (Lee et. al, 1997b). Figure 2 below shows a

graphical representation of Lee et. al (1997b) findings. The whole process creates an

upward trend. This surplus of production is balanced later when demand orders fall but this

swing dampens the firm, which is left with high costs and an inefficient supply chain, and

this in turn affects the industry as a whole (figure 2).

0

5

10

15

20

25

Consumer Sales

Retailer's Order

to Manufacturer

Wholesale's

Orders To

ManufacturerManufacturer's

Orders To

Supplier

Quan

tity

Weeks

Figure 2. Demand fluctuation along the chain Source: Lee et al. (1997)

Mitchell (1923) gives a good explanation of why the demand amplifies. He states the

retailer places an order but because of the shortage of supplies the manufacturers are only

able to fulfill 80% of the order. The firm could not produce to full capacity because there

was an unaccountable shortage of raw materials. The retailers become disappointed because

of the loss of profit associated with selling that 20%. In the next order they will increase

their ordering units to make up for the loss and to make sure they get what they ordered.

The retailers also order from other sources to make sure they have enough. However, when


143

the manufacturers sees an increase in the demand in the next order they increase the

demand order further to make sure they manufacture enough to cover the increasing

demand. The factory order to the supplier is even more to keep ‗safety stock‘ at hand due to

the unexpected demand increase. This whole process results in a huge surplus in the

demand orders that are not really required (Fisher, 1997).

Jay Forrester (1961), at Procter and Gamble, first observed demand amplifying as the

orders moved up the supply chain. He observed that consistent demand products like

Pamper (disposable diapers) with little customer demand variation produces high

fluctuation in the supply chain. With predicted birth rate, steady consumption rate and

stable sales, the higher variance in retailer order was reported. This information flow

(demand orders) was less variable with retailer‘s orders then compared to the distributors at

the factory. The orders to the 3M (suppliers) fluctuated ever more. At Hewlett-Packard‘s

printer supply chain the executive found the same distortion of orders from the resellers to

the company‘s circuit division as reported by Lee et. al (1997b).

Mitchell (1923) and Stalk and Hout (1990) reported the similar behavior of the supply

chain. In Barilla‘s pasta supply chain, Hammond‘s (1994) research shows higher

fluctuation in their weekly orders. Philips Electronics (De Kok et al. 2005) also reported the

presence of bullwhip.

Lee et al. (1997b) studied many supply chains and explains that the high degree of

uncertainty and variability causes the distortion in information of the demand orders that

lead every entity in the supply chain to stockpile. This results in operational inefficiency

and uses valuable working capital. The grocery industry as outlined by Efficient Customer

Response (ECR) estimated a $30 billion opportunity from their supply chain inefficiencies

(Kurt Salmon Associate, 1993). Manufacturers like Bristol-Myers Squibb, distributors such

as McKesson and retailers namely Longs Drug store in the pharmaceutical industry were

found to duplicate inventories. The same was found in the computer industry at Dell and

HP. The total chain contained more than one year‘s supply as reported by Lee et al. (1997).

This, then, is not the problem relating to firm alone but rather it is industry specific.

1.1 Bullwhip Across Different Industries

The affect of bullwhip in the industry thus affects the economy as a whole (Holmstrom,

1997). For this reason, the presence of bullwhip in the industry has been an interest for

many economists. Holt et al. (1968) found bullwhip in the TV industry, the jeans clothing

supply chain studied by Stalk and Hout (1990), automotive and camera (Blackburn 1993),

automotive (Avery et al. 1993; Taylor, 1999a), grocery (Holmsrom, 1997; Hammond,

1994; and Salmon, 1993), food by Lee et al (1995) and Fransoo & Woulters (2000), cement

industry (Ghali, 1987), retail industry (Blinder 1981; Gill & Abend, 1996; Mitchell, 1923),

pharmaceutical industry (Lee et a. 1997), cotton industry (Zymelman, 1965), home and

personal care (Lee et al. 1997a), toys (El-Beheiry et al. 2004), tobacco and tire (Fair, 1989),

electronic and furniture (Hejazi & Hilmola, 2004), machine-tool industry (Anderson et al.,

2000; McCullen and Towill, 2001), automobile industry (Blanchard, 1983; Allen, 1997)

and Cachon (2007) studied all industries and found the bullwhip effect.


144

1.2 Bullwhip In The Economy

Kondratieff (1984) studied the economy, discovering a phenomenon known as ―long wave‖

or Kondratieff cycle. Towill et al. (2006) relates the long wave to the bullwhip; the

overexpansion (demand amplification) is not profitable. Sterman (1986) explains how

‗over-expansion‘ in the economy leads to lower employment, production and cutting back

to eliminate what the industry had produced in the time of over-expansion and the overall

cost associated with it. This is similar to the bullwhip effect. In the time because of

increasing orders the firms sets up more equipment, raw material, employment but when

the surplus of production exists they need to cut back in order to sell the produced goods in

the inventory. Hence the disruptions and swings caused by bullwhip affects the economy as

it affects the industry and the firm itself. Metters (1997) pointed out that bullwhip results in

high costs due to the capacity on-cost and stock-out costs. The costs also increase because

of the demand stock-holdings has negative consequences. Below is a graphical

representation of the Kondratieff Wave (figure 3).

Figure 3. Periodicity in the economy Source: Kondratieff (1984)

1.3 Causes and Remedy

Andraski (1994) suggests that the bullwhip problem in the supply chain is 80% people

centered and 20% technology centered. Forrester (1980b) proposed that the problem lay in

the dynamically complex and non-linear nature of supply networks. Mason-Jones et. al

(2000) suggest the cause of bullwhip effect is ‗system induced and directly affected by

information and material flows and feedback loops‘. Senge & Sterman (1992) and Towill

(1982) related the bullwhip effect to the lack of managers‘ ability to think systemically.

Nienhaus et al. (2006) also believes that the causes of the bullwhip lie in human decision-

making.

1.3.1 Information Delay

Lee et al. (1997b) suggest that the demand distortion exists in the form of flow of

information. The driver here is the information and the decision is based on units ordered.


145

Lee et al. (1997b) also explain the distortion in information arising from demand forecast

updating, order batching, price fluctuation and rationing and shortage gaming.

1.3.2 Demand Forecast Updating

This occurs when the retailers base their order on the demand forecast rather than the true

market demand. As this information travels up the demand gets further amplified. Remedy:

Forrester suggested that eliminating intermediary like in the case of Apple and ―Dell

Direct‖ Lee et al. (1997b) explains how manufacturers such as IBM, HP and Apple request

sell-through data from their resellers so as to eliminate demand amplification.

1.3.3 Order batching

According to Lee et al. (1997a) is an end result of two factors, the periodic review process

and the processing cost. The manufacturers can reduce the information distortion by asking

for sell-through data so that they can manufacturers units based on sales than on the data.

Reducing transaction cost can help remedy the batch size and ordering cost. Nabisco, for

example, uses computer assisted ordering that is paperless reducing the ordering cost. This

system orders less and smaller batches which removes demand distortion and makes the

system more efficient.

1.3.4 Price fluctuations

These arise when the wholesale gives price discounts while the retailers do Every Day Low

Price (EDLP). These strategic buying causes buyers to capitalize on the discount offered on

the short period leaving the manufactures production schedule uneven along with the

inventory cost. Fisher (1997) demonstrated how ordering more to take advantage of pricing

strategy could be damaging. Manufacturers like Pillsbury, P&G and Kraft regularly carry

out trade promotions. Remedy suggested is for the manufacturer and buyer to sign a

contract under lower price or discount but the goods will be delivered over some period of

time. This way the inventory cost of the buyers and manufactures is reduced and also enjoy

strategic pricing policy.

1.3.5 Rationing and shortage gaming

This is when the manufacturers make decisions based on the order data. They need to

carefully understand when interpreting the orders. Gaming can take place when the retailers

order more to ‗self-protect‘ themselves when there is no shortage in reality but only

imagined as explained by Mitchell (1923). In a shortage situation manufacturers can

allocate supplies using a different rule Companies like HP, Texas Instruments, GM allocate

supplies proportionate to the market share of the retailer in the previous period.

Cancellation, and free returns also contribute to gaming. One remedy is to restrict buyer‘s

flexibility. Sharing forecast information is another way carried out by HP, Lands‘ End and

SUN.

1.4 Problems of Bullwhip

The problem of bullwhip effect is mainly two-folds: higher cost and lower efficiency.

Blinder (1986) suggests that there are ‗cost shocks‘ associated with the bullwhip as well.

Measurement also is a problem as there are many measure. The most popular measure has

also been criticized upon.


146

1.5 Measurement

Lee et al (1997a) used a model-based analysis to show the amplified data causes bullwhip

in three settings: correlated and seasonal demand, batch ordering and finite capacity. In a

study Cachon et al. (2007) the formula was used for the calculation of the bullwhip. A ratio

greater than one suggests the presence of demand amplification, while less than one means

that amplification does not exist in the supply chain.

When the seasonality is considered, Cachon et al. (2007) found variability dampening

manufacturing and retail industry then when seasonality was not considered. It was also

observed that seasonality did not play any part in the wholesale industry. The wholesalers

usually order large batch sizes because they are not restricted by space as the retailers.

Cachon et al. (2007) observed that bullwhip results are higher without seasonality

compared to seasonality. Variance ratio is most widely used measure and to detect bullwhip

effect. It is the ratio of the upstream and downstream demand variation. However, many

experts have stressed on the weakness of this measure.

1.6 Proposed Solutions

Many solutions have been proposed to eliminate bullwhip. Initially, Forrester (1958) and

Burbidge (1991) outline six supply chain solutions: control system, time compression,

information transparency, echelon elimination, synchronization principle and the multiplier

principle which was demonstrated by Wikner et al. (1992) and Anderson et al. (2000).

Wikner et al. (1991) of the operations manager school proposes remedies: ―improvement in

each echelons decision values, better timing of rules among different echelons, reduction of

time delays, removal of some distribution of echelons, better information flow along the

chain.‖

Lee et al. (1997a) also proposed some remedies: Information Sharing (IS) actions that

speed up information flow from downstream to upstream, Channel Alignment (CA) groups

aiming to increase co-ordination of process, Operational Efficiency (OE) increasing and

improving efficiency and reducing time. Miragliotta (2006) has provided a conclusive table

outlining all the remedies suggested by researchers around the IS-CA-OE paradigm and

also providing a new approach to eliminate bullwhip effect. Three modeling layers:

physical layer, reconstruction layer and control layer have been pointed out. The causes of

bullwhip fall in these layers. The determinants are the redesign action while the triggers are

the filtering actions. Each company faces different problems that cause the bullwhip hence

the delay in the physical layer is a trigger to analyze the firm‘s machines and process

capability etc (table 1).

Table 1. New taxonomy of the bullwhip

Determinants

(redesign actions)

Trigger (filtering actions)

Internal External

Physical layer Linear gain

Delays & feedbacks

Batching (quantity)

Machine

reliability

Process capability

Supply reliability

Reconstructor

layer

Delays + modeling

errors:

In accounting

In forecasting

Machine Errors:

In accounting

In forecasting

In measuring


147

Determinants

(redesign actions)

Trigger (filtering actions)

Internal External

In measuring

performances

performances

In accounting

Control layer Batching (quantity,

frequency)

Delays (in the

control model)

Irrational decision

making

Price

promotions/advertising

Campaigns

Shortage gaming

Financial deadlines

Source: Miragliotta, 2006

Geary et al. (2006) has summarizes approaches presented by different authors in the figure

below. The solutions are grouped as OR theory, filter theory, control theory, what-if

simulation, and ad-hocracy (figure 4).

Figure 4. Categorization of solutions of bullwhip Source: Geary et al. (2006)

1.7 Introduction to the various manufacturing systems

The demand amplification effect varies in different manufacturing paradigms. The degree

of amplification of fluctuation varies according to the manufacturing and supply chain

management policies implemented. Below the three policies are looked in greater detail.

1.7.1 Traditional

Mass production system is as old as industrialization itself. Henry Ford, founder of the Ford

Motor Company in 1913 manufactured cars that were simple, small and most importantly

cheap (Hounshell, 1984). The mass production system based its strength on economies of

scale. Producing Model T on a large bases with cheaper price lead him to capture the

market and sell the products at large. Lacey (1986) defines mass production as ‗single

purpose manufacturing combined with the smooth flow of materials in the assembly line;

large volume production, high wages and low prices.‘ Thus, economies of scale were

captured but it also created small delays very costly. The company moved towards

improving the assembly line to specialization and decentralization (Ford, 1923). When the

company decided to produce a new model, it required a change in the whole system from

Adhocracy

What-if Simulation

Control Theory

Filter Theory

OR Theory

Rea

l w

orl

d

pro

ble

m

Rea

l w

orl

d

So

luti

on


148

the suppliers, to the assembly. Pietrykowski (1995) explains how Ford and mass production

became popular but also lead to cul-de-sac situation (Raff & Trajtenberg, 1996). This was

because mass-manufacturing system had many weaknesses; most importantly, it was unable

to deal with rapid market changes. According to Duguay et. al (1997) mass production has

the following features: Reducing costs through economies of scale, production divided into

small, repetitive tasks, innovation as a means to improvement and existence of ‗adversarial‘

relationship with suppliers.

Duguay et al (1997) also explains how agile production is better as compared to mass

production. Although the era of technical changes had been well documented, little or no

quantitative analysis of mass manufacturing is available.

1.7.2 Lean

After the 1973 oil shock the focus of Japanese companies shifted to reducing or eliminating

wastes thereby cutting costs (Monden, 1983). The lean strategy came from the realization

of the scarcity of resources and a fiercely competitive industry. Lean methodology

implemented the strategic leveraging of resources.

Taichi Ohno, an engineer at Toyota Motors, claims that the lean revolution did not happen

overnight, however, it was a process of small step-by-step improvements that took thirty

years (Ohno, 1988). Lean processes started primarily at Toyota under two engineers, Taichi

Ohno and Shigeo Shingo. Later, the two authors published books on production processes

and so the lean revolution first began in Japan and spread to the rest of the world.

There were huge gaps of efficiency between Toyota and other car manufacturers; this

solidified the superiority of lean philosophy. Based on Dyer‘s (1994) research the following

table shows the efficiency at Toyota in comparison to other car manufacturers. Toyota

implemented a lean framework while the rest were grounded in the traditional supply chain

management and production methods (table 2).

Table 2. Car manufacturers and efficiency compared

Variable Chrysler Ford General

Motors Nissan Toyota

Quality 148.4 120.9 131.7 110.6 78.6

Cycle Time 80.7 77.7 86.5 51.4 50.0

Auto Inventory/Sales 9.8% 8.4% 8.1% 5.1% 2.3%

Suppl. Inventory/Sales 10.7% 10.1% 11.2% 8.8% 7.3%

Auto Maker ROA 4.1% 4.2% 2.8% 5.5% 13.0%

Suppliers ROA 5.4% 5.5% 4.8% 5.6% 7.1% Source: Dyer, 1994

Lean manufacturing was ‗an enhancement of mass production‘ (Motwani, 2003). After

mass manufacturing and the assembly line continuous improvements were being sought

after in production and the supply chain. Toyota sought improvement in every process; the

system that emerged was called the ‗lean‘ system. This lean methodology focused on

cutting costs and reducing/eliminating the seven lean wastes. The lean approach divided

activities into three types; value-adding activities, non-value adding activities and necessary

non-value adding activities. The seven lean wastes were identified: Transport, inventory,

motion, waiting, over processing, overproduction and defects.


149

Basically, the lean approach stressed on continuous improvement, minimum defects, and

efficiency by producing in small batches. The major shift among mass production system

and lean is that lean system emphasizes the importance of constant innovation within and

without the company. On the other side, mass production just stresses producing large

quantities to benefit from economies of scale. Lean emphasized strategy; doing more with

less; and adapting to the continuously changing environment whilst reducing costs as much

as possible.

The improvement in efficiency of manufacturing from mass to lean production was evident;

companies outside the car industry also started applying lean principles. Also, in the

beginning lean was an organization-wide philosophy, however, it soon incorporated

networks of suppliers becoming an inter-organizational approach.

Lean system proposed Shojinka, meeting variability in demand through flexibility. This

reflects the fact that demand variations were accepted as a fact and production was changed

to meet the demand and avoid excess inventories (Monden, 1983: 99).

According to Knill (1999) successful implementation of lean requires focus on five parts:

Supplier programs, continuous improvements, flexibility, eliminate waste and zero waste.

As the figure 4 shows lean‘s strength lay in connecting the strategic to the operational level,

thus, having an excellent strategic implementation.

1.7.3 Agile

The concept of agility owes its birth to the field of social systems. Emery and Trist (1965)

proposed the notion that organizations were open systems that existed in environments.

They divided uncertainties within environments in four types, the most intense which was

termed turbulent field; where uncertainty arose from the environment itself. Figure 5 shows

the changing variables that an organization has to deal with constantly.

Figure 5. Strategic and operational levels Source: Hines et al. (2004)


150

In business, the concept of agility first came forward from the Iacocca Institute in 1991

(Nagel & Dove, 1992). Agility was defined as the ‗ability to respond with ease to

unexpected but anticipated events‘ (Oleson, 1998).

Another definition of agility in manufacturing was by Sarkis (2001):

Agile manufacturing = flexible manufacturing + lean manufacturing

Narasimhan et al. (2006) wrote that agile processes were focused on superior flexibility and

speed of delivery and that they were ‗built upon‘ lean methodology. Now agile is being

termed as an alternative to or improvement of‘ lean manufacturing (Mason-Jones et al.,

2000). Lean manufacturing is adopted where there is a stable demand while agile is suitable

to markets with volatile demand. Mason-Jones et al. (2000) also introduced the leagile

paradigm where both lean and agile processes can be used by the utilization of a

‗decoupling point‘.

Parkinson (1999) claims that the agility revolution in the manufacturing came about largely

due to the increasing intensity of competition. Gunasekaran (1998) claims that focus

towards customer satisfaction, business system integration, shortening product life cycles,

customized services, and flexibility in meeting volatile demand gave rise to what is known

as agile manufacturing. This variety-variability matrix gives Christopher‘s (2000)

perspective on agile and lean systems. This differentiation between agile and lean is

essential in how both the systems cope with demand fluctuations (figure 6).

Figure 6. Organization and its environment

Source: Yauch (2011)

Figure 7. Variety-variability matrix

Source: Christopher (2000)


151

Many definition of agility proposed however all are synonymous to responsiveness to

changing market requirements (Ramesh & Devadasan, 2007). Thus, there is a focus on

speed in the agile paradigm.

2. Methodology

Data was collected using case studies. All data is based on research already conducted by

academic professionals. As the study is qualitative, the causes of the demand amplification

effect and the measures implemented to counteract the effect were analyzed and there

effectiveness assessed.

Initially, it was planned to include consistent cases for the ease of comparability. However,

the cases on bullwhip are still limited, especially, in terms of manufacturing paradigms and

there was also a lack of consistency. Due to this, eight cases were included that had

significant information for them to be analyzed in detail. For example, Forrester (1958) and

Zymelman‘s (1965) studies were synthetic – a simulation ran on industry data. However,

most of the cases are reflections of actual companies and the results of implementation of

new policies. Where it was not explicitly mentioned if the company was lean, agile or

traditional Narasimhan et al.‘s (2006) variables were used to differentiate between

manufacturing paradigms. Here, low performers were equated to Mass manufacturers. The

cases were divided on the bases of purchasing and inventory policy, production methods,

priorities (whether minimizing cost or customer service) into three categories: Traditional,

lean and agile.

A significant weakness of this case analysis was the difference of external variables such as

industry type, firm type, geographic location, demand patterns and size. In retrospect, these

differences might give a deeper look into the demand amplification effect.

Primarily, the goal of the research is to conclude the effect of different manufacturing

paradigms on the bullwhip effect. Going deeper, it aims to throw light on the causes of

those differences. Unfortunately, only one case was discovered that covered agile

manufacturing in detail. Below is a list of the cases included in the study (table 3).

Table 3. Case studies in analysis

Author Industry Number

of Echelons

Manufacturing

paradigm Details

Forrester, 1958 Hard goods

(simulation)

3 Traditional Simulated

Zymelman, 1965 Cotton Textile 3 Traditional Simulated

Hejazi & Hilmola,

2006.

Electronic

supply chain

3 Traditional Make-to-stock

Hammond, 1995 Barilla SpA,

Pasta Supply

Chain

3 Traditional Inefficient Supply

Chain

Taylor, 2000 Automotive

Components

3 Lean JIT and pricing

policies, minimum

stock.


152

2.1 CASE 1: Automotive Components (Taylor, 2000)

2.1.1 Discussion

The supply chain analyzed here was a 3-echelon chain. The automotive steel supply chain

included a steel producer, two steel service centers, six component producers and many

automobiles OEM (original equipment manufacturers). The figure below shows the supply

chain with the companies involved (figure 7).

Figure 7. The Steel Automotive Supply Chain

Source: Taylor (2000)

Author Industry Number

of Echelons

Manufacturing

paradigm Details

Kok et. al, 2005 Philips

Electronics

3 Lean -

Hejazi & Hilmola,

2006

Furniture

Supply Chain

3 Lean Assemble-to-order

Fransoo &

Woulters, 2000

Food 3 Lean Highly seasonal

demand

McCullen &

Towill, 2001

Mechanical

Engineering

Products

3 Agile JIT, SMED, delivery

time given most

importance

Steel Mill Automobile

OEM‘s

First Tier Component Producers

Steel Service Centres

Steel Mill

Steel & Alloy

British Steel

Distribution

Albion Pressed Metal

Tallent Engineering

LDV

GKN Sankey

Krupp Camford

Wagon Group


153

A disaggregated approach to the measurement of demand amplification was used where

individual products through the value stream were analyzed. This approach is more specific

than mapping demand amplification across many product lines. The project was based on

the demand of two high volume parts demanded by car manufacturers. To quantify demand

amplification, demand and activity data were needed. Demand data was based on actual

demand and forecasts, while activity data included production and dispatch activity

gathered from various points in the value chain.

Figure 8 shows that the OEM demand is stable, however, as the bullwhip is transferred

across the chain Service Centre Advanced Orders to Mill are greatly amplified. Some

causes were identified for the bullwhip effect were demand variability, supply variability,

functional silos and decision-making, stock minimization and J.I.T and pricing policy.

0

20

40

60

80

100

120

1 3 5 7 9 11 13 15 17 19 21 23 25 27

Service Center weekly

call to mill

OEM Demand

To

nn

es

Weeks 1-27 Year 1998

Figure 8. Variability along the whole chain Source: Taylor (2000)

The paper then sought to educate personnel about the adverse effects of demand

amplification. Then, a demand management team was formed. This team comprised of one

person from each company. This was an effort to centralize decision making of demand

across the supply chain so as to reduce the amplification of the bullwhip as stressed in the

literature.

The demand management team after understanding the bullwhip made some decisions:

1. Weekly demand between the component manufacturer and the service center

were set at the average demand from the OEM – 14 tonnes/week.

2. Weekly demand between the service center and steel mill was set at one

coil/week for weeks.

3. Demand information from the OEM was shared directly with the component

manufacturer, service center and steel mill through the demand management

team.

4. Significant changes in the demand on OEM would be directly communicated

through the demand management team.

5. Safety stocks were determined at different point along the chain

6. Bulk order prices were made available for smaller but regular batches.


154

2.1.2 Results

After implementation of, primarily, a central demand management team the fluctuations of

mill dispatches have been significantly reduced, even when the OEM demand fluctuation is

greater than in the last observation.

The OEM demand during the first ten weeks of the project remained at 13.5 tonnes/week as

anticipated (figure 9). Demand regularization reduced any amplification that might have

travelled along the chain. There was some unexpected variation in demand, which was

coped by the system, because of the centralized decision making of demand. The figure

above also shows the steady reduction in inventory after implementing the aforementioned.

These findings encouraged the senior management to carry on with the study. The delivery

from service center to component manufacturer improved to 100% on time delivery – an

increase from 95%.

0

10

20

30

40

50

60

1 3 5 7 9 11 13 15 17 19 21 23

Tonnes

Mill Despatch to

Service Center

Component

Manufacturer call-in to

Service Center

OEM Actual Demand

Weeks

Figure 9. Actual demand and supply during the trial Source: Taylor (2000)

Taylor noted that the demand fluctuation was dependent on decision of the amount to order

from upstream suppliers and the previous order in queue. If less stock were ordered

previously the decision to order extra inventory for the next period would be the initial

trigger of amplification.

Managerial decision-making was on a functional basis whereby managers made logical

decisions according to the best of their knowledge, however, these decision were beneficial

for the company in question and not the whole supply chain. It was seen that these

managers who decided inventory and production levels thought only about their own

responsibilities and not the wider system. Thus, unintended consequences were realized

elsewhere in the supply chain eventually resulting in the further amplification.

The implementation of a lean strategy throughout the supply chain reduced many small

fluctuations. These fluctuations were damped primarily because of demand smoothing and

a level scheduling policy.


155

Also, inventory levels were kept to a minimum as a lean policy stresses on minimum

inventory and Just-in-time. It was seen that companies expected their suppliers to meet the

variations in demand through JIT ‗no matter how variable demand was‘. In the steel supply

chain, buffer stock was not kept in consideration of demand variability, these factors further

amplified the demand. Furthermore, demand was also amplified as steel mills delivered a

minimum order of up to 70 tonnes batches; however, the OEM demand was constant at

13.5 tonnes per week. Minimum inventory policy of the steel mill created unpredictable and

haphazard demand.

The demand management team reduced Forrester‘s (1958) information distortion through

the chain. The three steps that were implemented proved reduction in the demand

amplification in 13 weeks. An increase in delivery reliability was also noted which was

attributed to the decrease in demand variability. The project stated that even in a JIT

system, safety or buffer stock acted to reduce the amplification.

2.2 CASE 2: Cotton Textile (Zymelman, 1965)

In this case study, Zymelman (1965) incorporated research from the cotton textile gray

goods industry through actual prices, production, mill inventories, unfilled orders and

personal interviews. Then, a stabilization policy was simulated and its effects documented.

The U.S. cotton gray goods industry at that time was in a market structure known as perfect

competition, where there were many competitors involved in the market. The market was

subject to periodic fluctuations. However, the negative effects of fluctuations were realized,

such as, swing in workforce demand, prices and subsequently profits were noted.

Zymelman (1965) proposed that collaboration between the industry actors would help

stabilize these inconsistencies. The following diagram shows the supply chain layout and

the number of firms involved.

Figure 10. The cotton supply chain Source: Zymelman (1965)

Mills produced cotton while converters process that into cut cotton cloth for the firms,

which sew them into garments. It was noted that an accumulation of inventories at the mill

significantly reduced the bargaining power of the mills, thus, converter companies were

essentially price-takers. Converters companies based their buying decision on the demand

for cotton cloth; the price expectations from the mills and the converter‘s inventory needs.

It took three months for a producer to deliver an order to the converter and the producers

delivered from a buffer stock as it took six month for a producer to process an order.

=Material Flow

=Information Flow

Many Firms 300 Mills 1200 Converters


156

The simulation was run on an analog computer as flexibility and its timesaving properties

were given an emphasis over accuracy and processing capacity of a digital computer. This

research was based on analyzing a small variation in demand that resulted in fluctuations of

inventory levels, production levels and prices. The trigger of amplification was an external

variable i.e. the demand from cutter companies.

2.2.1Results

The simulation showed that the demand from cutters was not the only trigger of the

fluctuations. Changes in unfilled orders and inventories were directly related to the level of

fluctuations. Zymelman‘s (1965) study proposed that the decision of production at the mills

should also be based on the ratio of unfilled orders to inventories. It was also discovered

that inventories and unfilled orders were to be given more importance in order to reduce the

fluctuations in inventories. Secondly, it was concluded that the companies should base

decision-making on the ratio of unfilled orders to inventories.

In the LEAP program (automotive components) endogenous variables were also found to

impact amplification. According to Zymelman, only variables that arose from outside the

system (such as demand) caused fluctuations in inventory, production and prices.

2.3 CASE 3: Household Appliances (Forrester, 1958)

Jay Forrester ran a simulation on the household appliances industry. It was a mass

manufacturing system with three tiers in the supply chain. This study was the first of its

kind, giving breakthrough insights into the bullwhip effect in 1958.

Forrester identifies flows of information, materials, money, manpower, and capital

equipment as amplifiers of fluctuations. In a mass production network, the demand

amplification is greater than among lean network. The action suggested by Forrester were

faster order handling, better forecasting of demand, and better sales data.

Figure 11 shows the effect of a sudden 10% increase in orders. Whereas the rise in order is

10% the subsequent rise in factory output is by 40%. Moreover, there is also a lag of

several weeks. Figure 12 shows the effect of 10% increase in sales but with ‗reasonably

reliable sales data‘. Here, the rise in orders only affects a rise of 16% in the factory output.

This study stresses how better sales data could reduce the bullwhip effect.

0

500

1000

1500

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Retailers Order From Customers

Weeks

Qu

an

itty

Figure 11. Response of production-distribution system to a sudden 10% increase in

retail sales Source: Forrester (1958)


157

0

200

400

600

800

1000

1200

1400

1 4 7 10 13 16 19 22 25 28 31 34 37 40

Retailers' Orders From Custom…

Weeks

Quantit

Figure 12. Orders and production output comparison Source: Forrester (1958)

2.4 CASE 4: Electronics And Furniture (Hejazi & Hilmola, 2006)

The authors take into account two cases. The first is the analysis of the furniture supply

chain in Finland. The furniture company utilizes a make-to-stock strategy. The second case

is of an electronic manufacturer. However, this electronic manufacturer utilized an

assemble-to-order type of policy. They followed a ‗pull‘ demand management system,

which can be categorized as lean. This difference among strategies also gives some

variation of results and deeper insights into using a different manufacturing strategy.

These differences were because of the nature of the companies or supply chain. The

electronic manufacturer had a high product variety while a lower volume, and the opposite

was true for the furniture company, as shown in the diagram.

2.4.1 Furniture Supply Chain

The furniture company implemented a make-to-stock supply chain strategy in the local

finish furniture industry. The have expanded aggressively in the past through Greenfield

investments. In the year 2002, they had annual revenue of 43 million euros. The supply

chain observed consisted one product family that included dressers, commodes and chests.

This was a three tier supply chain comprising of manufacturer, distribution center and

retailers (figure 13). Figure 14 displays existence of a seasonal pattern in demand. Figure

15 show a comparison of sales of retailers, purchase orders from distributor and deliveries

from the manufacturer.

Figure 13. The furniture supply chain Source: authors

Manufacturer Distribution Centre Retailer


158

0

50

100

150

200

250

1 2 3 4 5 6 7 8 9 10 11 12

2001

2002Qu

an

tity

Months

Figure 14. Seasonal demand pattern Source: Hejazi & Hilmola (2006)

0

50

100

150

200

250

300

350

1 7 13 19 25 31 37 43 49 3 9

Sales

Deliveries

Weeks 2002- Beginning of 2003

Qu

an

tity

Figure 15. Sales and deliveries comparison

Source: Hejazi & Hilmola (2006)

Order batching was identified as the major cause of bullwhip effect in this company.

The cause of order batching was the production lot sizes of the manufacturer; a minimum

quantity of order was needed to produce. Another reason was long set-up times

(of 15-45 minutes) when producing a different order. Manufacturing was also constrained

by a bottleneck; the painting, varnishing and staining of products. Thus, manufacturing lot

sizes were directly connected to the amplification of demand.

An implementation of a leaner policy would reduce set up times, which would mean

smaller batches of production. Together, these two would increase the efficiency of

production and decrease lead times.


159

2.4.2 Electronics Supply Chain

The electronics company is involved in assemble-to-order and is a part of a large

international supply chain. Over the past, they have grown organically and in 2002 there

annual revenue was at 54 million euros. It was a three-tier supply chain consisting of

distributor, manufacturer and system integrator. The product chosen for the study was a

platform used for transmitting cable signal. The demand mostly was regular but there were

times of erratic demand (figure 16). Figure 17 shows the demand versus time. Again, there

is a seasonal fluctuation in demand. Figure 18 shows the existence of demand amplification

across the chain.

Figure 16. Electronics supply chain Source: Hejazi & Hilmola (2006)

0

200

400

600

800

1 11 21 31 41 51 9 19 29 39 49

Sales Orders

Weeks 2000-2001

Qu

an

ti

Figure 17. Sales orders for optical node Source: Hejazi & Hilmola (2006)

Figure 18. Three tier demand comparison Source: Hejazi & Hilmola (2006)


160

Two facts can be concluded from the above graph; first, the difference order quantity and

production quantity, and second, the difference between order and supply quantity. Note the

difference of intensity and time lag between demand and manufacturing.

2.4.3 Discussion

Both of these companies were evolving towards becoming a fully lean approach based

firms. However, as the electronics company follows an assemble-to-order production it has

evolved more towards becoming lean comparatively, as assemble-to-order policy reflects a

pull system.

The furniture company is relatively less lean as it follows a make-to-stock. Also constraints

and bottlenecks also limited the productive capacity of furniture business. There is a clear

difference among bullwhip of the two companies, which is attributable to the degree they

have evolved at being lean.

The authors claim that production lot sizes and transfer batches were partly able to reduce

the bullwhip. They identify order batching (size of manufacturing batches) as the main

cause of amplification. They also discovered that there was a difference between production

quantity, order quantity and supply quantity. These variables combined to contribute to the

bullwhip. Orders were received in large quantities while deliveries were made in smaller

batches. In both cases, thus, order batching, was the major cause of amplification for the

upstream of the supply chain. VMI (vendor managed inventory) was proposed as the

solution for the furniture company. It was also suggested reducing set up times and

delivering products in more frequent and smaller batches would reduce the effect for the

electronics chain.

2.5 CASE 5: Glosuch Supply Chain (McCullen & Towill, 2001)

McCullen and Towill (2001) incorporate both experiential and simulation techniques to

study the bullwhip effect in the precision mechanical engineering industry. They analyze

Glosuch supply chain‘s six years time-series data to evaluate pattern in demand and supply

through the chain.

This supply chain consists of three echelons comprising of a U.K. factory, a U.K. finished

goods warehouse and a number of overseas warehouses (figure 19).

Figure 19. The glosuch supply chain Source: McCullen & Towill (2001)

Firms in the study implemented Total Quality Management (TQM), partnering across the

supply chain, Single Minute Exchange of Dies (SMED), small and more frequent delivery

batches, and JIT production. Trucks picked supplies from docks every hour and delivered to

customer‘s plants. However, even in such a time-sensitive supply chain manufacturer‘s

excess capacity was to blame subsequent amplifications. The entire chain is forecast driven

and all echelons have there own decision-making.

Factory (UK) Warehouse (UK) International Warehouse


161

0

50

100

150

200

250

300

18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75

Production

SalesQ

uan

tity

Weeks Figure 20. Production and sales comparison

Source: McCullen & Towill (2001)

An analysis of the original Glosuch supply chain revealed that overseas subsidiaries greatly

exaggerated demand and safety stock were depleted because of such demand. After

experiencing amplification because of greatly magnified sudden demand by foreign

subsidiaries it was decided to implement an agile strategy. The company implemented a

rapid response manufacturing program and information system integration through the

supply chain.

The effect of implementing this agile strategy was reflected in the reduction of information

and material processing lead-time from 23 weeks to 2 weeks. Evidently, a migration from

lean to agile policy reduced lead times by 91%. As part of the study, three inquiries were

made: Was there a bullwhip effect in the Glosuch original supply, was it reduced by

implementing an agile strategy and does agility reduce variability?

Figure 20 shows how bullwhip was reduced in the original supply chain and reduced to a

greater degree after implementing an agile strategy. The graph also shows steady reduction

in variability as Glosuch climbs the learning curve. An average bullwhip reduction of 36%

was realized in two echelons and in a period of four years global inventory was reduced by

45%. Thus, rapid response manufacturing and information system integration had achieved

the results that were predicted in literature.The figure shows the graph of both production

and sales. It is clearly evident how amplification is slowly reducing with time as the system

learns. Table 4 shows the eventual reduction in Glosuch‘s global inventory. In year 4, there

is a total reduction of 45% which is almost half.

The authors attempt to implement an approach that eliminates the initial trigger of

amplification at the source. Their solutions were highly effective. It also shows that the

greatest improvement in terms of bullwhip reduction is when a company evolves from lean

to agile.

Table 4. Reduction in global inventory

Year 0 Year 1 Year 2 Year 3 Year 4

Weeks 31 26 22 20 17

Difference of Stock - -5 -4 -2 -3

Total % change - -17% -29% -35% -45% Source: McCullen & Towill (2001)


162

2.6 CASE 6: Food (Fransoo & Wouters, 2000)

Two three-tier supply chains were analyzed in a lean environment; one was meals product

and the other salads. In this study, the author tested whether the use of real time EPOS data

by the manufacturer would make an impact on the bullwhip effect. Daily demand for both

of the supply chain was analyzed for 72 days. The main factors of demand variability were

promotional activities, price fluctuations and weather patterns. The weather factor was

especially evident in salads. Figure 21 shows the supply chain. Table 5 shows the

amplification ratios of both the supply chains. The results show that there is greater

amplification in supply chain of greater fluctuating demand

Figure 21. The food supply chain Source: Fransoo & Wouters (2000)

2.6.1 Results

According to Table 5, the bullwhip effect in salads is far more than in meals this is due to

sudden increases in temperature, which affected the demand for salads. Authors claimed

EPOS (electronic point-of-sale) information to the supply would reduce the bullwhip.

Another variable that could stabilize demand was a more stable production. The sharing of

EPOS data signifies to an extent the sharing of real-time demand information, which, in

effect reduces the information distortion

Table 5. Amplification ratios in salad and meals supply chain

Echelons Bullwhip Effect meals Bullwhip Effect Salads

Production 1.75 1.23

Distribution Center 1.26 2.73

Retail Franchisee 1.67 2.09

Source: Fransoo & Wouters (2000)

2.7 CASE 7: Philips Electronics (Kok et al, 2005)

In 1999, Philips Semiconductor and Philips Optical Storage formally studied the bullwhip

effect in their supply chain. These companies operate under Philips Electronics, a global

consumer electronics manufacturer. This study highlights the actions and results of the

attempt made by Philips to reduce demand amplification. Increased fluctuations were

observed in Philips Semiconductor, which was a supplier of Philips Optical Storage.

Collaborative Planning Project was undertaken with the objective of reducing fluctuations.

The process of wafer-fabrication was identified as a bottleneck.

Supplier A

Supplier B

Retail Franchisee Retailer

Distribution Centre


163

The Collaborative-Planning (CP) Project was implemented with the ultimate aim of

reducing the bullwhip. It centralized decision-making of production and deliveries to the

CP staff. The figure 22 shows the reduction of the ratio between on-hand stock and work-

in-process from 70% to 30%. The results of centralizing decision-making were overall

stock reduction, reduction of obsolescence and increased agility in the chain. Financially,

these reflected as a 1.5% increment of profit. Furthermore, they freed a capital of $5 million

that could be used to fulfill working capital requirements. These results support the fact that

central demand management greatly diminishes the bullwhip.

0

20

40

60

80

100

Ratio Between On-

Hand stock and

Work-in-Process

Per

cen

t

Week Number Figure 22. Ratio between on-hand stock and work-in-process

Source: Kok et al. (2005)

2.8 CASE 8: Barilla Spa (Hammond, 1994)

Barilla Spa is a pasta company in Italy. After inefficacies in the supply chain a JIT program

was applied to its supply chain in collaboration with the company‘s distributor, Cortese. A

pilot implementation of JITD (Just-In-Time Distribution), between Barilla and its

distributors reduced the stock-out rate from 5% to less than 1%. This was achieved through

daily demand information sharing through fax or the Internet. After this success, a full JITD

program was implemented. Below is a graph of the reduction in stock after implementation

of the JITD. The co-ordinates for the following two graphs were not available. Therefore, it

was taken as it is from the case (figure 23). There was also an obvious reduction in

percentage stock outs from 7% to virtually none. This improvement is a direct effect of JIT

distribution, which makes stock management more efficient. Figure 24 shows the bullwhip

reduction before and after the implementation of the new policy.

Figure 23. Stock out rate from marchese DC to retailers, before and after the start of JIT

Source: Hammond (1994)


164

Figure 24. Bullwhip reduction in barilla supply chain

Source: Hammond (1994)

Not only did these improvements effected efficiency but there was also a rise in customer

service noticed, as orders were delivered on schedule. Small and more frequent deliveries

also influenced the reduction of amplification. As the paper showed promising results,

Barilla plans to incorporate the JIT distribution in their manufacturing processes.

3. RECOMMENDATIONS

In pursuing the reduction of demand amplification, the companies in the supply chain

should be thoroughly analyzed. This is because demand amplification can result from a

multitude of factors. Then, all that information should be synthesized by the use of

techniques such as rich-picture mapping, cause-and-effect diagrams, input-output diagrams

and value chain analysis among others. This step is necessary to view the supply chain as a

system. This is because bullwhip is fundamentally a systemic phenomenon and it can only

be realized if the whole system is scrutinized. Subsequently, the supply chain members

need to collaborate to counter this phenomenon by relevant measures along the chain. The

study has expressed that a reduction and near elimination is possible. It, however, also

depends on the strategic priorities of all companies in the supply chain. Every firm is in a

stage where it has to start by treating one of the causes of bullwhip. That option is the major

amplifier of demand. The company should then implement policies step by step to diminish

the bullwhip altogether.

McCullen and Towill (2001) recommend firms to implement their four principles in order

to systemically address all the root causes of the bullwhip effect. These principles were so

effective that Geary et al.‘s (2006) solution were a derivative of these.

CONCLUSION

After a brief look at table 6, the effectiveness of central demand management is evident.

This is because centrally managing demand targets three root causes of amplification –

namely, information delay, demand forecast updating and, rationing and shortage gaming.

Fluctuations amplify, as the upstream member of the supply chain is not aware of the actual

end-consumer demand, however, sharing demand data diminishes such possibility. Thus,

central demand management has been identified as the major solution to the bullwhip effect

with a high certainty of dampening the effect.


165

Table 6. Analysis of the cases

Author Industry

Manufac-

turing

Paradigm

Cause Solution Improvement/

Results

Zymelman,

1965

Cotton

Textile

(simulation)

Traditional Volatile demand,

inefficient inventory

policy, speculation

of demand

Production

schedule should

be based on the

ratio of unfilled

orders over

inventory

Reduction in

fluctuation after

applying solution to

the model

Forrester,

1958

Hard goods

(simulation)

Traditional Simulation with

changes in variables

such as reliability of

sales data

More reliable

sales data, better

order handling,

Decrease of factory

production variation

from 40% to 16%,

thus amplification

reduced

Hejazi &

Hilmola,

2006.

Electronic

supply chain

Traditional Large single

purchase quantities

set by the distributor

Combined

purchases, Sales

order not equal to

manufacturing lot

size

Prevented

amplification‘ and

‗Partly avoided the

bullwhip‘ (Hejazi &

Hilmola, 2006)

Hammond,

1994

Pasta Supply

Chain

Traditional General

inefficiencies in the

supply chain

JIT Distribution,

flexible delivery

load, advanced

demand

management and

marketing

integration,

Reduction in stock

outs from 6% to

near 0%.

46.7% decrease in

inventory at

Marchese DC.

Significant

reduction in

bullwhip.

Fransoo &

Woulters,

2000

Food Lean Rationing and

shortage gaming.

Shared EPOS

(electronic point-

of-sale)

information

Results not

published

Hejazi &

Hilmola,

2006

Furniture

Supply

Chain

Lean Order Batching:

Large

manufacturing lot

sizes of the

upstream member

Shorter set-up

times, Vendor

Managed

Inventory

‗Prevented

amplification‘ and

‗Partly avoided the

bullwhip‘ (Hejazi &

Hilmola, 2006)

Kok et al,

2005

Electronics

(Philips)

Lean Long lead times,

wafer-fabrication as

a bottleneck,

shortening product

life cycles and

volatile demand,

Inter-

organizational

collaboration in

demand

management

Overall stock

reduction, reduction

of obsolescence and

increased agility.

1.5% increases in

profit.

40% decrease in

ratio between

on-hand stock and

work-in-process.


166

In conclusion, this report proposes that collaboration between supply chain members as a

necessary ingredient in the fight against demand amplification. Collaboration among

members of the supply chain also reduces the effect as more steps are taken in order to

think about the whole system rather than one company. Collaboration has been identified as

a major counter-measure for it treats most of the problems simultaneously by facilitating

information exchange. The functional silo decision-making concept is thus becoming

extinct as companies realize the potential of holistic thinking.

Also, bullwhips that are caused by order batching can only be partly reduced by measures

such as small but regularized deliveries and vendor-managed inventory.

The delivery policies were also a major cause for inducing bullwhip. For example, in the

automotive supply chain (Taylor, 2000) minimum batches sizes for order were identified as

main cause. This was countered by an agreement where ‗smaller but more regularized‘

order would be delivered at the same cost (Taylor, 2000).

In terms of the bullwhip, there is a continuum upon which manufacturing paradigms have

evolved. This has largely been a result of the change in the macro-environment. Table 6

shows the how natural selection has made traditional manufacturing rare because of a

general focus towards efficiency and market responsiveness.

As agile and lean developed after the identification of the demand amplification, they are

better equipped to deal with demand amplification. Lean philosophy stresses efficiency and

cost effectiveness; JIT distribution and SMED methods enable companies to respond to

fluctuations in demand much better as compared to traditional production (table 7).

Author Industry

Manufac-

turing

Paradigm

Cause Solution Improvement/

Results

Taylor,

2000

Automotive

Components

Lean Functional silo

decision-making,

JIT expectations

from the next

upstream partner,

pricing policy,

minimum order size.

Central demand

management team,

removal of price

policy on small

but regular orders,

regular schedules

Obvious reduction

in amplification,

better customer

service, decrease in

inventory levels

McCullen

& Towill,

2001

Mechanical

Engineering

Products

Agile Safe stock

depletion, order

exaggerations,

unresponsive supply

chain

(lead time=23

weeks)

Pull system

introduced,

partnership

arrangements,

planning cycle

changed from

monthly to daily

Reduction of lead-

time by 91%.

Average bullwhip

reduction of 36%.

45% reduction in

global inventory in

4 years.


167

Table 7. Evolution of bullwhip

Environmental

Factors

Industrialization

and Post-War

Reconstruction

Free Market

Economy And Cut

Throat Competition

Faster Economic

Cycles and Shorter

New Product

Development/

Product life cycles

Manufacturing

Paradigm

Continuum

Traditional Lean Agile

Primary Focus Make product

available to wider

public

Price reduction

based on

economies of

scale

Cost Efficiency

Reliability

Market

Responsiveness

Customer

Satisfaction

Affect

on Bullwhip

Demand

amplification

identified

Reduction in the

bullwhip effect

Near total reduction.

Agile goes a step further; the main focus of agility is market responsiveness. This means

that agile supply chain is quickly able to respond to changes in the marketplace. For

instance, fashion clothing industry is uncertain demand with new fashions coming every

few months. For such companies the primary aim is to the get product to the customer as

soon as possible. In such a case, agile manufacturing is the best possible solution.

However, there is a trade-off of going from lean to agile – higher costs.

It is stressed that there is a need for further research on fast and agile supply chains in order

to understand, in greater depth, how agility counters the bullwhip, and to codify the

methods that are more effective in dealing with the bullwhip across supply chain.

References

Allen, D.S. (1997). Do inventories moderate fluctuations in output? Federal Reserve Bank

of St. Louis Rev. 79 (4), 39–50.

Anderson, E., Fine, C., & Parker. G. (2000). Upstream volatility in the supply chain: The

machine tool industry as a case study. Journal of Production and Operations

Management, 9(3) 239-261.

Andraski, J.C. (1994). Foundations for a successful continuous replenishment programme.

International Journal of Logistics Management, 5 (1), 1–8.


168

Avery, C.N., Lariviere, M.A., & Harrison, J.M. (1993). Case S-OIT-IC at the Raychem

Corporation, Graduate School of Business, Stanford University, CA.

Batchelor, R. (1994). Henry Ford, mass production, modernism and design. Manchester:

Manchester University Press.

Blackburn, J.D. (1991). Time-Based Competition: How time based competition is reshaping

global markets. BusinessOne Irwin, Homewood

Blanchard, O.J. (1983). The production and inventory behavior of the American automobile

industry. Journal of Political Economy, 91(3) 365–400.

Blinder, A. (1981). Retail inventory investment and business fluctuation. Brookings Papers

on Economic Activity, 2, 443-505.

Blinder, A. (1986). Can the production smoothing model of inventory behavior be saved?

Quarterly Journal of Economics, 101(3), 431-53.

Burbidge, J.L. (1991). Period batch control (PBC) with GT – the way forward from MRP,

Paper presented at the BPCIS Annual Conference, Birmingham.

Cachon, T., Randall, G., & Schmidt, G.M. 2007. In search of the bullwhip effect. Journal of

Manufacturing & Service Operations Management, 9(4) 457–479.

Chase, R.B., Aquilano, N.J., & Jacobs, F.R. (2001). Operations Management for

Competitive Advantage. McGraw Hill/Irwin: Boston

Chen, F., Drezner, Z., Ryan, J.K., & Simchi-Levi, D. (2000). Quantifying the Bullwhip

Effect in a simple suly chain: The impact of forecasting, lead times and information.

Journal of Management Science, 46(3), 436–443.

Christopher, M., (2000). The Agile Suly Chain: Competing in Volatile Markets. Journal of

Industrial Marketing Management, 29(1), 37-44.

De Kok, T., Janssen, F., van Doremalen, J., van Wachem, E., Clerkx, M., & Peeters, W.

(2005). Philips Electronics synchronizes its suly chain to end the bullwhip effect.

Interfaces, 35(1), 37–48.

Duguay, C. R., Landry, S., & Pasin, F. (1997). From mass production to flexible/agile

production. International Journal of Operations and Production Management, 17(12),

1183-1195.

Duguay, C., Landry, S., & Pasin, F. (1997). From mass production to flexible/agile

production, International Journal of Operations and Production Management, 17(12),

1183-1195.

Dyer, J. (1994). Dedicated Assets: Japan‘s Manufacturing Edge, Harvard Business Review,

72 November/December

El-Beheiry, M., Wong, C.Y. and El-Kharbotly, A., (2004). Empirical quantification of

bullwhip effect (with application to a supply chain). Proceedings 13th Working

Seminar on Production Economics, IGLS, vol. 3, 83–95.

Emery, F., & Trist, E. (1965). The causal texture of organizational environments. Journal

of Human Relations, 18, 21- 32.

Fair, R.C. (1989). The production-smoothing model is alive and well. Journal of Monetary

Economics. 24, 353–370.

Fisher, M. (1994). National Bicycle. The Wharton School Case, Philadelphia, PA.

Fisher, M. (1997). What is the right supply chain for your product? Harvard Business

Review, March/April, 105–116.

Ford, H. (1923). My Life and Work. Garden City, NY: Garden City Publishing Company.

Forrester, J. (1961). Industrial Dynamics. MIT Press and John Wiley & Sons: New York.

Forrester, J. W. (1980). System Dynamics -- Future Opportunities. 7-21. In: Augusto A.

Legasto, Jay W. Forrester, and James M. Lyneis, eds. System Dynamics. TIMS Series

in the Management Sciences, 14. Amsterdam: North Holland.


169

Fransoo, J.C., & Wouters, M.J.F. (2000). Measuring the Bullwhip Effect in the supply

chain. Supply Chain Management: An International Journal, 5(2), 78–89.

Geary, S., Disney, S.M., & Towill, D.R. (2006). On Bullwhip in Supply Chains historical

review, present practice and expected future impact. International Journal of

Production Economics. 101(1), 2-18.

Ghali, M., (1987). Seasonality, aggregation and the testing of the production-smoothing

hypothesis. American Economic Review. 77(3). . 464-469

Gill, P. & Abend, J. (1996). Wal-Mart: the supply chain heavyweight champ, Supply Chain

Management Review, 1, 8-16.

Gunasekaran, A. (1998). Agile Manufacturing: enablers and an implementation framework.

International Journal Of Production Research, 36(5), 1223-1247.

Hammond, J. (1990). Quick Response in the Apparel Industry. Harvard Business School

Case: Cambridge, MA.

Hammond, J. (1994). Barilla SpA (a). Harvard Business School case #9-694-046.

Hejazi, A., & Hilmola, O. (2004). Manufacturing lot sizing as a source of Bullwhip

Effect—case study of electronics and furniture supply chains, Proceedings of the

Thirteenth Working Seminar on Production Economics, 4, 125-140.

Hines, P., Holweg, M., & Rich, N. (2004). Learning to evolve: A review of contemporary

lean thinking, International Journal of Operations & Production Management,

24(10), 994 – 1011.

Holmstrom, J. (1997). Product range management: A case study of supply chain operations

in the European grocery industry. International Journal of Supply Chain Management,

2(8), 107–115.

Hounshell, D. A. (1984). From the American system to mass production. Baltimore: The

John Hopkins University Press.

Hout, T., & Stalk, G. (1990). How time-based management measures performance,

Strategy & Leadership, 18(6), 26-29.

Knill, B. (1999). How lean manufacturing matchs todays business. Material Handling

Engineering. October, 87-91.

Kondratieff, N.D. (1984). The Long Wave Cycle. Richardson and Snyder: New York.

Kurt Salmon Associates. (1993). ECR: Enhancing the Consumer Value in the Grocery

Industry. Food Marketing Institute: Washington.

Lacey, R. (1986). Ford: The men and the machine. Little Brown & Co.: n/a

Lee, H.L., Padmanabhan, V., & Whang, S. (1997a). Information distortion in a supply

chain: the bullwhip effect. Journal of Management Science. 43(4), 546–558.

Lee, H.L., Padmanabhan, V., & Whang, S. (1997b). The bullwhip effect in supply chains.

Sloan Management Review, 38(4), 93–102.

Lee, H.L., Padmanabhan, V., & Whang, S. (1995). The Paralyzing curse of Bullwhip Effect

in a Supply Chain. Stanford University, March 17th

Lee, H.L., So, K.C., & Tang, C.S. (2000). The value of information sharing in a two-level

supply chain. Journal Of Management Science, 46(5), 626–643.

Mason-Jones, R., Naylor, B., & Towill, D. (2000). Lean, agile or leagile? Matching your

supply chain to the marketplace. International Journal of Production Research,

38(17), 4061-4070.

McCullen, P., & Towill, D. (2001). Practical ways of reducing bullwhip: the case of the

Glosuch global supply chain. Control, December/January, 24–30.

Metters, R. (1997). Quantifying the bullwhip effect in supply chains. Journal of Operations

Management, 15(2), 89–100.


170

Miragliotta, G. (2006). Layers and mechanisms: A new taxonomy for the Bullwhip Effect.

International Journal Production Economics, 104 (2), 365–381.

Mitchell, T. (1923). Competitive illusion as a cause of business cycles. Quarterly Journal of

Economics, 38(4), 631–652.

Monden, Y. (1983). Toyota Production System. Atlanta, Ga: Institute of Industrial

Engineers.

Motwani, J. (2003). A business process change framework for examining lean

manufacturing: A case. Industrial Management and Data Systems, 103 (5/6), 339.

Nagel, R.N., & Dove, R. (1992). 21st Century Manufacturing Enterprise Strategy, Iacocca

Institute, Lehigh University, Bethlehem, PA.

Narasimhan, R., Swink, M., & Kim, S.W. (2006). Disentangling leanness and agility: An

empirical investigation. Journal of Operations Management, 24(5), 440-445

Nienhaus, J., Ziegenbein, A., & Schönsleben, P. (2006). How human behaviour amplifies

the bullwhip effect—A study based on the beer distribution game online, Journal of

Production, Planning & Control, 17(6), 547–557.

Ohno, T. (1988). Toyota Production System – Beyond Large-scale Production. Productivity

Press: New York.

Oleson, J. (1998). Pathways to agility: Mass customization in action. John Wiley & Sons:

New York.

Parkinson, S. (1999). Agile manufacturing, Work Study, 48(4), 134-137.

Pietrykowski, B. (1995). Fordism at Ford: Spatial Decentralization and Labor Segmentation

at the Ford Motor Company. Journal of Economic Geography, 71(4), 383-401.

Raff, D., & Tajtenberg, M. (1996). Quality-adjusted prices for the American automobile

industry: (1906-1940). In: Bresnahan, T., & Gordon, R. (1996). The Economics of

New Goods. University of Chicago Press: Chicago, Ch: 2.

Ramesh, G., & Devadasan, S.R. (2007). Literature review on the agile manufacturing

criteria. Journal of Manufacturing Technology Management, 18(2), 182-201.

Rudberg, M., & Olhager, J. (2003). Manufacturing Networks and Supply Chains: An

Operations Strategy Perspective. Omega: The International Journal of Management

Science, 31(1), 29-39.

Sarkis, J. (2001). Benchmarking for agility, Benchmarking. An International Journal, 8(2),

88-107.

Senge, P.M., & Sterman, J.D. (1992). System thinking and organisational learning: Acting

locally and thinking globally in the organisation of the future. European Journal of

Operational Research, 59(1), 137–145.

Stalk, G.H., & Hout, T.M. (1990). Competing Against Time, How Time Based Competition

is Reshaping Global Markets. Free Press: New York.

Sterman, J.D. (1986). The Economic Long Wave, Theory and Evidence. Systems Dynamics

Review, 2 (Summer), 87–125.

Sterman, J.D. (1989). Modeling managerial behavior: Misperceptions of feedback in a

dynamic decision making experiment. Journal of Management Science, 35(3),

321–339.

Taylor, D.H. (1999). Measurement and analysis of demand amplification across the suly

chain. International Journal of Logistics Management, 10(2), 55–70.

Tayler, D.H. (2000). Demand Amplification: has it got us beat? International Journal of

Physical Distribution and Logistics Management, 30(6), 515-533.

Towill, D.R. (1982). Dynamic analysis of an inventory and order based production control

system. International Journal of Production Research, 20(6), 671–687.


171

Wikner, J., Naim, M.M., & Towill, D.R. (1992). The system simplification approach in

understanding the dynamic behaviour of a manufacturing supply chain. Journal of

Systems Engineering, 2(2), 164–178.

Wikner, J., Towill, D.R. & Naim, M. (1991). Smoothing supply chain dynamics.

International Journal of Production Economics, 22(3), 231–248.

Yauch, C. (2011). Measuring Agility as a Performance Outcome. Journal of Manufacturing

Technology Management, 22(3), 384-404.

Zymelman, M. (1965). A Stabilization Policy For Cotton Textile Cycle, Journal of

Management Science, 11(5), 572-580.

Date post:	01-Apr-2021
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

The Bullwhip Effect in Different Manufacturing Paradigm: An...

Documents